Towards Automatically Detecting Whether Student Is in Flow

Abstract

Csikszentmihalyi\'s flow theory states the components (e. g., balance between skill and challenge) that lead to an optimal state (referred to as flow state, or under flow experience) of intrinsic motivation and personal experience. Recent research has begun to validate the claims stated by the theory and extend the provided statements to the design of pedagogical interactions. To incorporate the theory in a design, automatic detector of flow is required. However, little attention has been drawn to this filed, and the detection of flow is currently still dominated by using surveys. Hence, within this paper, we present an automated detector which is able to identify the students that are in flow. This detector is developed using a step regression approach, with data collected from college students learning linear algebra from a step-based tutoring system.